Open-source local voice assistant avoids cloud API calls
Athena is a voice-to-voice AI assistant that runs on local hardware instead of sending speech to . It combines for , Qwen3.5-397B for generating replies, Orpheus 3B for spoken output, and SNAC for audio handling. The runtime is written in C++ and uses a four-part process, with no Python needed during normal use.
One optional is used only once during setup to convert an emotion model offline. Athena can speak with emotional sounds such as laughs, sighs, and gasps. It can read basic emotion in a person’s voice, keep and personality across conversations, and handle long .
If the user talks over it, it stops speaking and keeps what it already said in the conversation context. It is designed to run without , telemetry, or API keys on a consumer GPU plus system RAM.
Key points
- Athena runs as a fully local voice assistant with no cloud service required.
- It includes , language generation, voice output, audio handling, memory, and interruption handling.
- It keeps and personality between sessions.
- It uses C++ at runtime and avoids Python during normal operation.
- No API keys or telemetry means fewer external and no usage-based API charges.